Annual Report Department

2.8Logistics management

2.8.1Reverse logistics

Reverse Logistics (RL) studies how systems for the collection, separation, and treatment of products at end-of-life can efficiently support reuse and recycling. The study focusses on three topics. First, location and process selection models are developed. Using facility location models, an analysis is made of how product recovery affects reverse logistics network design in various supply chain contexts. To support the recycling companies in the selection and planning of the recycling processes, a decision support tool is developed based on linear programming. The second topic deals with the impact of reverse logistics on vehicle routing. Using probabilistic analysis and continuous approximation techniques, the relative efficiency of various vehicle routing approaches is modelled and analysed, including various approaches to integrate the collection with the distribution activities and various postponement strategies. To support municipal waste collection, linear programming and column generation models for the sectorisation problem are developed and combined with a probabilistic approach to analyse the effects of source separation, visit frequency, and other parameters on the fixed and variable transportation costs. Topic three deals with new local search heuristics for node, arc and general routing problems. These routing problems form the basis for many RL collection problems. It is shown that the approaches applied for node routing problems, in particular k-Opt and Guided Local Search, are also successful for arc and general routing problems.

Publications and reports: WP2001-20, WP2003-12, WP2003-19, WP2003-16, WP2003-22

Scientific staff: P. Beullens, D. Cattrysse, D. Van Oudheusden

2.8.2Arc routing and districting

The term arc routing refers to routing problems where the key service activity is to traverse arcs of a transportation network rather than to visit a number of clients concentrated in discrete nodal points (node routing). Typical arc routing applications include refuse collection, postal delivery, winter gritting, road maintenance, etc. Both mathematical models and software tools to support these activities are much less developed than for the nodal counterpart. In this research we focus on the following topics: new algorithms for arc routing, districting, and location models for arc routing.

• 
  New algorithms for arc routing. The well-known k-Opt approach for node routing is suitably adapted to deal with arcs or edges. Based on this technique new local search algorithms are designed to solve the General Routing Problem and the Capacitated Arc Routing Problem. The procedures turn out to be very effective and systematically identify optimal or near optimal solutions on large instances.
  
• 
  Districting for arc routing. District design is an intermediate step between the location of the facilities and the planning of the routes. The problem involves the partitioning of a road network into subnetworks or districts such that each district contains a centrally located facility from where the (arc) routing activities are organised within the district borders. Important district design guidelines to allow efficient routing are derived and incorporated in several algorithms. The procedures are tested in large graphs derived from the road network in Flanders.
  
• 
  Location-arc routing and location-districting models. In this study we integrate location decisions with routing and / or district design. Polynomial and pseudo-polynomial procedures are developed to solve location–capacitated arc routing problems on simple networks when demand splitting is allowed. Complexity issues are also studied.
  

Publications and reports: WP2001-20, WP2002-09, WP2003-35, 03DC02

Scientific staff: D. Cattrysse, L. Muyldermans, D. Van Oudheusden

2.8.3Strategic logistics management through rational transport mode choice

This research work originates from an ever more relevant society theme, namely the mobility problem. Because of a global integration tendency and some demographic and socio-cultural evolutions, the last decades the number of movements worldwide has increased enormously. Transport infrastructure capacity is at several spots insufficient.

The company transport modal choice for its freight flows is elaborated in order to enable companies to deal, in a sustainable way, with the mobility problems in the different transport modes, not neglecting at the same time broader logistic tendencies. Further the integration of the transport function in general logistics management is aimed and an interaction between inventory management and transport management is built up in order to help companies to reach global logistic optima to stay competitive in a global market. Centralization versus decentralization in distribution is a related research topic. Decision support tools, based on Multi Criteria Analysis and Total Logistic Cost Calculation, are elaborated in order to objectify logistics decisions in general and transportation choices in particular.

Related projects: Promodi

Publications and reports: WP2003-01

Scientific staff: L. Pintelon, B. Vannieuwenhuyse

2.8.4Management of process changes

The subject of this research is the management of process changes. Management, to invest in production capabilities, especially in environments with complex production processes, frequently changing production technologies and large production volumes, often uses the introduction of process changes. The normative lecture literature presents few models that give structural insight in process improvement policies for manufacturing. Moreover, in these models, the effect of the decision is deterministic or decisions are proposed under deterministic conditions. Management however deals with different sources of uncertainty while planning for process changes. Therefore this research focuses on policies for changing the manufacturing process recipe under different types of uncertainty.

Publications and reports: WP2003-09

Scientific staff: L. Gelders, L. Pintelon, J. Vits

2.8.5Inventory pooling of repairable spare parts

Inventory pooling is a type of inter-company co-operation in which companies share their inventories. Co-operation usually takes the form of lateral transhipment from a location with a surplus of on-hand inventory to a location that faces a stockout. This co-operation is an effective way to reduce a company’s inventory cost without negatively impacting its performance. This research focuses on the inventory pooling of repairable spare parts, and more specifically on applications in the airline industry. Previous research in the area of inventory pooling has mainly focused on co-operation among ‘bases’ which are assumed to be owned by a single ‘parent-firm’ whose objective is to maximise some measure of aggregate performance. Little research has been done to capture the story of inventory pooling when the co-operating companies are independent and interested in maximising their own performance rather than that of the entire system. The objective of this research is to develop models for designing and evaluating various pooling mechanisms amongst independent companies.

Publications and reports: WP2003-05, WP2003-29

Scientific staff: D. Cattrysse, D. Van Oudheusden, F. Van Utterbeeck, H. Wong

2.8.6Maintenance concept optimization

This research addresses the problem of designing optimum, i.e. efficient and effective, maintenance concepts. A maintenance concept is defined as the set of maintenance interventions of various types (reactive, preventive, predictive, proactive, opportunistic,…) and the general structure in which these interventions are foreseen. While optimising maintenance policies is well covered in literature, optimizing maintenance concepts is far less exposed in literature. Literature describes generic concepts like TPM (total productive maintenance), RCM (reliability centered maintenance), BCM (business centered maintenance),… No research so far tackles the problem of designing a customized concept taking into account the specific situation on-hand in a real-life industrial situation.

The aim of this research is to focus on this issue and to develop a decision framework for building a customized industrial maintenance concept. The framework also addresses the policy optimization question and provides a decision scheme for choosing the right models for optimizing the selected policies within the concept developed.

Publications and reports: WP2003-35, WP2003-36, WP2003-37, WP2003-17

Scientific staff: G. Waeyenbergh, L. Pintelon

2.8.7Maintenance auditing and performance systems

Auditing, performance reporting, benchmarking,… are hot issues for every management discipline. The purpose of this research is to design a method which goes beyond the classical performance reporting methods ; hereby focusing on integration in the business vision (the so-called balanced scorecard approach) and modelling the maintenance activities. Important elements are establishing targets for performance indicators and industrial benchmarking. The research addresses different business settings and corresponding strategies in manufacturing industry and

service industry.

Related projects: UZ Leuven (TD)

Scientific staff: L. Pintelon

2.8.8Link between business and maintenance strategy

The aim of this research is to develop a framework for choosing appropriate maintenance strategies. These maintenance strategies concern decisions on structural and infrastructural elements of maintenance. These decisions are taken on the highest management level and from the basis for further tactical and operational decisions.

Although maintenance is an important factor in the overall manufacturing strategy, the relationship between busines, manufacturing and maintenance strategy is rarely addressed in literature. However, research in this area is needed and also expected by industry.

This research starts out with establishing the state-of-the-art on this topic in Belgian industry. Next a framework for formulating a maintenance strategy will be designed based on the well-known framework of Hayes & Wheelwright. A DSS using AHP principles will be developed to operationalize this theoretical framework.

Related projects: Laborelec

Publications and reports: WP2003-06, WP2003-31, WP2003-28

Scientific staff: L. Pintelon, K. Pinjala

2.8.9Object/objective oriented maintenance and asset management

The fundamental principle in this research is that maintenance and asset management should be approached in an integrated way. This means that besides technological considerations, also arguments concerning information and knowledge management as well as economic and managerial aspects should be taken into account.

The aim of this research is to develop (quantitative and qualitative) models and concepts in order to support the maintenance and asset management decision process. This should lead to new theoretical insights and useful practical guidelines to improve management performance in this area.

Related projects: Sidmar - Arcelor, Philips Lighting

Scientific staff: G. Zhu, L. Pintelon, L. Gelders

2.8.10Collaborative supply and demand networks

Nowadays, companies operate in a globalizing world under constantly increasing competition. They should first recognize the difference between running the business today and running the business tomorrow. This study will focus on surviving tomorrow, or long-term survival. To hold a competitive advantage, working together or collaboration will be crucial.

At present, the supply chain is often considered as an isolated system. However, in reality, it has many interconnections. The system can be looked at as a supply web in which the supply chain is a subsystem. A structured supply web is called a supply network. Two types of collaboration can be distinguished, namely vertical and horizontal collaboration. Vertical collaboration can be defined as coordination between parties that perform complementary activities and/or services. Horizontal collaboration can be defined as the cooperation between parties that perform the same type of activities and/or services.

The overall research question is formulated as follows: “How should companies collaborate in a network (vertically and horizontally) to improve the cost/benefit ratio based upon the characteristics of each firm.” The answers to this question will provide input for a decision support framework for the implementations of collaboration in a networked organization.

Publications and reports: WP2003-39, WP2003-24

Scientific staff: L. Gelders, L. Pintelon, K. Naesens

2.8.11Operational urban police research

Organising the local police function represents a substantial part of the budget of Belgian cities. Operational activities, such as the patrolling and intervention function, contribute highly to this cost because they are labor-intensive and are organised around the clock. Local police authorities therefore want to utilise these resources as optimal as possible, to deliver an adequate service to the community. The authorities lack however descriptive models to evaluate adequately their system performance and to assess the quality and implications of proposed policy changes. At the same time they ask for prescriptive systems that guide their decisions and that direct the management of the patrol and intervention function. In this research we focus on the following topics: patrol deployment, sector design and dispatch procedures.

• 
  Patrol deployment: what is the effect of the followed patrol strategy on relevant performance measures, and how should it be adapted to increase the delivered service or to achieve the same performance with less resources?
  
• 
  Sector design: to manage activities more efficiently authorities split their police zone in smaller geographical units (sectors), to which patrol units are assigned. Confronted with congestion effects, the design of sectors highly influences the achievable performance. To assist this sectorisation decision authorities need both descriptive and prescriptive models.
  
• 
  Dispatch procedures: which unit should be dispatched if a call for service arrives. This choice influences not only the current reaction time, it possibly also influences subsequent reaction times as unavailabilities of patrol units build up.
  

Publications and reports: WP2003-33, WP2003-24

Scientific staff: M. Moonen, D. Cattrysse

2.8.12Outsourcing

Almost all business journals carry articles on how and why to outsource. There is also an abundance of conferences and websites on outsourcing. This wealth of information seems often difficult to digest. Too often in practice such an outsourcing project turns out to be not successful. The aim of this research is to explore the outsourcing decision making process and to develop a generic framework for supporting decisions. Different levels of outsourcing are distinguished: from operational subcontracting to strategic partnerships and collaborative networks. This framework is validated in various fields: logistics, maintenance, facility management,… This will also allow to extend the framework with features to incorporate the specific requirements for each of these situations.

Related projects: Faceo

Scientific staff: L. Pintelon, B. Vannieuwenhuyse

2.8.13Internal logistics

This research focusses on the optimization of internal logistic processes, such as material handling, warehousing, packaging,… This includes both the organization of the physical flows and the corresponding information flows. Typical problems addressed here are warehousing layout, order picking, logistic performance measurement,… The tool kit used consists of mathematical algorithms and simulation as well as organizational optimization procedures.

Related projects: NMBS, Belgian Post

Publications and reports: WP2003-24

Scientific staff: L. Gelders, L. Pintelon